• Journal of KIISE:Software and Applications
• /
• v.35 no.11
• /
• pp.692-700
• /
• 2008

In some previous works on musical genre classification, human experts specify segments of a song for extracting musical features. Although this approach might contribute to performance enhancement, it requires manual intervention and thus can not be easily applied to new incoming songs. To extract musical features without the manual intervention, most of recent researches on music genre classification extract features from a pre-determined part of a song (for example, 30 seconds after initial 30 seconds), which may cause loss of accuracy. In this paper, in order to alleviate the accuracy problem, we propose a new method, which extracts features from representative segments (or main theme part) identified by structure analysis of music piece. The proposed method detects segments with repeated melody in a song and selects representative ones among them by considering their positions and energies. Experimental results show that the proposed method significantly improve the accuracy compared to the approach using a pre-determined part.

• Spatial Information Research
• /
• v.10 no.2
• /
• pp.331-343
• /
• 2002

Recently, as various fields of applications using space-borne imagery have been emphasized, interests on integrated analysis or fusion using multi-sources are also increasing. In this study, to investigate applicability of multiple imageries for further regional-scaled application, DN value analysis and multi-resolution classification by using KOMPSAT EOC imagery and Landsat 7 ETM+image data in the Namyangju-city area were performed, and then this classified results were compared to land-use thematic data at the same area. In case of classified results by using muff-resolution image data, it is shown that linear-type features can be easily extracted. furthermore, it is expected that multi-resolution classified image can be effectively utilized to urban environment analysis, according to results of similar pattern by comparative study based on multi-buffered zone analysis or so-called distance analysis along main road features in the study area.

• The KIPS Transactions:PartB
• /
• v.8B no.6
• /
• pp.657-668
• /
• 2001

It is difficult that web documents are classified with exact user intention because existing document classification systems are based on word frequency number using single keyword. To improve this defect, first, we use keyword, a query, domain knowledge. Like explanation based learning, first, query is analyzed with knowledge based information and then structured user intention information is extracted. We use this intention tree in the course of existing word frequency number based document classification as user information and constraints. Thus, we can classify web documents with more exact user intention. In classifying document, structured user intention information is helpful to keep more documents and information which can be lost in the system using single keyword information. Our hybrid approach integrating user intention information with existing statistics and probability method is more efficient to decide direction and range of document category than existing word frequency approach.

• Journal of KIISE:Software and Applications
• /
• v.31 no.4
• /
• pp.525-536
• /
• 2004

Gene expression profile is numerical data of gene expression level from organism, measured on the microarray. Generally, each specific tissue indicates different expression levels in related genes, so that we can classify disease with gene expression profile. Because all genes are not related to disease, it is needed to select related genes that is called feature selection, and it is needed to classify selected genes properly. This paper Proposes GA based method for searching optimal ensemble of feature-classifier pairs that are composed with seven feature selection methods based on correlation, similarity, and information theory, and six representative classifiers. In experimental results with leave-one-out cross validation on two gene expression Profiles related to cancers, we can find ensembles that produce much superior to all individual feature-classifier fairs for Lymphoma dataset and Colon dataset.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.26 no.5
• /
• pp.409-415
• /
• 2016

Recently several techniques for disaster prevention based on IoT(Internet of Things) are being developed. In this paper, a new smart pattern classification method for electric loads is proposed. CT(Current Transformer) data are extracted from electric loads, and then the sampled CT data are converted using FFT and MFCC. FFT and FMCC data are used for the input data of neural networks. Experiments were conducted using FFT and MFCC data for 7 kinds of electric loads. Experiments results indicate the superiority of MFCC in comparison to FFT.

• Proceedings of the Korean Society for Emotion and Sensibility Conference
• /
• 2009.05a
• /
• pp.209-210
• /
• 2009

수면단계는 수면감성을 평가하는데 있어서 중요한 생리지표로서 사용되어왔다. 그러나 수면다원검사를 이용한 전통적 수면단계 분류방법은 뇌전도, 안전도, 심전도, 근전도 등을 종합적으로 측정하므로 수면단계를 비교적 정확히 분류할 수 있지만 피험자에게 심한 구속감을 주는 문제가 있다. 본 연구에서는, 각성상태에서 교감신경계가 지배적인 반면에 수면 중에는 부교감신경계가 더 활동적인 점에 착안하여 수면단계를 간단히 분류할 수 있는 방법을 찾고자 수면단계에 따른 심박동변이도(heart rate variability: HRV)를 분석하였다. 단일채널 심전도를 이용하여 수면단계별로 HRV 의 교감신경계/부교감신경계 활성도의 비율을 분석한 결과, W(wakefulness) 단계가 NREN(non REM) 2 단계, 3 단계, 4 단계에 비하여 높게 나타났으며, NREM 4 단계는 REM(rapid eye movement) 단계와 NREM 1단계에 비하여 낮게 나타났다. 또한 교감신경계/부교감신경계 활성도 비율의 수면단계에 따라 변화하는 양상은 W, REM, NREM 1, 2, 3, 4 단계의 순으로 단조 감소하였다.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.24 no.12B
• /
• pp.2447-2454
• /
• 1999

This paper proposes a N-APPEM(Nonlinear A Posteriori Probability Estimation Method) with a frame normalization method to conventional classification network to increase speech recognition ability. It also tests the recognition ability of the classification and prediction neural networks for the Korean isolated digits. From the experimental results, the prediction network with MLP(Multi-Layer Perceptron) achieves the highest recognition ability of 98.0%. The prediction requires very complicated networks increased linearly with the number of incoming speech categories. However, the classification network with the N-APPEM and the normalization improves the recognition ability up to 85.5% with a sin81e network, which is almost 12.0% improvement.

• Smart Media Journal
• /
• v.12 no.2
• /
• pp.36-41
• /
• 2023

With the advent of applicant tracking systems, online recruitment has become more popular, and recruitment fraud has become a serious problem. This research aims to develop a reliable model to detect recruitment fraud in online recruitment environments to reduce cost losses and enhance privacy. The main contribution of this paper is to provide an automated methodology that leverages insights gained from exploratory analysis of data to distinguish which job postings are fraudulent and which are legitimate. Using EMSCAD, a recruitment fraud dataset provided by Kaggle, we trained and evaluated various single-classifier and ensemble-classifier-based machine learning models, and found that the ensemble classifier, the random forest classifier, performed best with an accuracy of 98.67% and an F1 score of 0.81.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.33 no.6
• /
• pp.1067-1076
• /
• 2023

Recently, as variant malware has increased, the scale of cyber hacking incidents is expanding. To respond to intelligent cyberhacking attack, machine learning-based research is actively underway to effectively classify malware families. However, existing classification models have problems where performance deteriorates when the dataset is obfuscated or sparse. In this paper, we propose a hybrid dataset that combines features extracted from ASM files and BYTES files, and evaluate classification performance using FNN. As a result of the experiment, the proposed method showed performance improvement of about 4% compared to a single dataset, and in particular, performance improvement of about 30% for rare families.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.11 no.4
• /
• pp.1203-1209
• /
• 2010

LED lights as a newly-growing industry are highly energy-efficient and have drawn lots of attention due to higher illuminance and longer life compared to other light sources. In this study, MR16 lights with one high-power LED were considered for the replacement of the previous halogen lights or LED lamps with many LEDs. Thus, fundamental experiments of LED on a MCPCB, 3-dimensional numerical analysis for heatsink design and performance tests of the prototype lights have been done and the MR16 LED lights have been successfully developed.